implementation experience report

Although I still haven't seen FIXptr, I've suggested a few times that the 
only parts of XPointer that most developers need in a Web context are the 
bare ID and /1/2/3.

After looking back over my previous development with XLinkFilter 
(http://www.simonstl.com/projects/xlinkfilter/), I'm quite happy to see 
that the implementation I provided - in the filter and in sample code - got 
along quite nicely with tree-based identifiers (equiv to /1/2/3) and bare 
IDs.

The LocationFilter 
(http://www.simonstl.com/projects/xlinkfilter/code/com/simonstl/sax/location/LocationFilter.java) 
just maintained a stack to give XLinkFilter a means of expressing link 
origins for inline links using the TEI-based XPointer syntax then available.

The image map demonstration 
(http://www.simonstl.com/projects/xlinkfilter/code/linkmap.htm) I've been 
using for my presentations relies on IDs for its simple functionality.

I don't believe that this obsolete bit of code demonstrates that additional 
functionality is unnecessary.  I would suggest, however, that it 
demonstrates that it's possible to do an enormous amount with very little.

(And if I could write the code to support it, so could a _lot_ of other 
people.)

Simon St.Laurent - Associate Editor, O'Reilly & Associates
XML Elements of Style / XML: A Primer, 2nd Ed.
XHTML: Migrating Toward XML
http://www.simonstl.com - XML essays and books

Received on Saturday, 5 May 2001 14:47:26 UTC